Log in

Core Concepts

The three shared assets (Skills, Knowledge Graph, Conversations), the workspace they live in, and the records that keep them honest.

Headways is a small platform built around one job: make AI use compound across a team instead of staying trapped in individual chats.

There are three shared assets that compound, one workspace surface they live in, and three records that keep the whole thing honest.

The Three Shared Assets

Shared Skills

Reusable workflows the team authors and runs together. Pre-connected to your stack. Versioned, reviewed, owned.

Knowledge Graph

Your org's context — playbooks, decisions, customer history — kept fresh and accessible to every workflow.

Shared Conversations

The thread, the work product, the sources. Move them between teammates with one click.

The Surface

Agent Workspace

The desktop app where prepared work, reviews, and conversations all converge into one Inbox.

Connectors

Plug the workspace into the tools your team already uses — Slack, Drive, Confluence, Jira, Notion, SF, HubSpot, Gmail, Loom.

The Records

Output Ledger

Every AI output tracked: what was produced, accepted, edited, or discarded.

Governance

Protection rules, review queues, and audit trail across every Skill and Connector.

Daily Library Review

The quality loop that promotes winning Skills and retires underperformers.

Runtime

Runtime Adapters

Run the same Skill in the workspace, Claude Code, Cursor, Copilot, or Codex.

Analytics

Skill adoption, usage, and outcomes by team.

How They Compose

The pattern is the same for every piece of work the team puts through Headways:

  1. A trigger fires. An email lands, a Slack thread starts, a ticket changes, a meeting hits the calendar. A Connector saw it.
  2. A Skill runs. A versioned, owned Skill drafts the output, drawing on the Knowledge Graph for context.
  3. A conversation forms. The thread, the draft, and the sources land together in the Agent Workspace as a Shared Conversation.
  4. The owner reviews. Approve, edit, send back. The decision goes into the Output Ledger and through any applicable Governance gates.
  5. The good stuff is promoted. Conversations that produced great work get their patterns lifted into the Knowledge Graph or a new shared Skill, so the next team picks up higher.

That's the loop. Three assets, one surface, three records — composed into a workspace where AI use compounds across the team.